Building Scalable
Robust Backends
Architecting, scaling, and modernizing high-growth platforms.
Specialized in Ruby on Rails, Payments, Distributed Systems, and AI-Assisted Engineering.
About Me
I am a Senior Software Engineer with 13+ years of experience building and modernizing backend systems, most of them with Ruby on Rails. Today I work on payments and subscriptions at Scribd, where correctness and reliability matter more than speed.
I have led teams and platforms across fintech, edtech, e-commerce, and messaging. My focus is event-driven architecture, performance, and upgrading legacy codebases without downtime.
I build AI-assisted engineering workflows: agent skills, MCP servers, and an LLM wiki that turn incident work into searchable knowledge. One of them helped cut a production Sentry backlog by 48%.
I am highly collaborative. I like pair programming, mentoring other engineers, and explaining technical trade-offs to non-technical stakeholders.
Above all, I care about writing code that makes a difference and about leaving systems easier to run than I found them.
Technical Arsenal
Backend
Payments
Databases
Architecture
AI Engineering
DevOps
Testing
Frontend
Tools
Social
Professional Experience
Senior Software Engineer
Digital Reading · Subscriptions · Payments
- ▹Cut payments Sentry backlog 48% (202 to 105 issues) and event volume 42% (250k to 144k) through an AI-assisted weekly triage and 74 closed tickets.
- ▹Eliminated 229k Stripe webhook errors for 10.8k users and fixed checkout 500s for 2,537 by persisting each subscription's Stripe account and handling detached payment methods.
- ▹Closed a payments log-leak audit across 60+ files, redacting gateway credentials, tokens, and PII, and wrote a rotation runbook covering all 10 payment vendors.
- ▹Retired legacy payment code by deleting 6,000+ lines (Bango, Braintree Forward API, Fortumo, fraud scoring) and queued 7 unused tables (7M+ rows) for production drop.
- ▹Migrated 13k retired-carrier subscribers into standard dunning and cancellation using a dry-run-first admin command with dedupe and currency-mismatch rollback.
- ▹Moved payments alerting to Terraform: adopted 121 Datadog monitors through a batch-import script and repaired or deleted 9 dead monitors.
- ▹Stabilized Google Play Billing by closing 13 tickets and adding missed-renewal reconciliation, stuck-order tooling, and a v2 payload state resolver.
- ▹Automated weekly Sentry triage with Claude Code skills that classify issues, group root causes, and draft Jira tickets for approval, shared with the payments team.
- ▹Built an LLM wiki and a 21-tool MCP server that enforces its schema, capturing 85 incident investigations, 49 plans, and 33 post-deploy verifications.
- ▹Authored a Nix devenv and agent guidance for the monolith: AGENTS.md, 8 routed guides covering 17 code patterns, and a linter for the docs.
Technical Lead
EdTech · K-12 Communications
- ▹Architected a distributed event-driven messaging platform serving 1M+ district users by integrating Ruby on Rails, Apache Kafka, and resilient background services.
- ▹Scaled SMS delivery throughput from 1,200/min to 20,000/min (16x boost) by implementing a Kafka pipeline with Karafka and optimized batch dispatch.
- ▹Modernized a critical legacy Rails monolith to Rails 7 with zero downtime, eliminating message bottlenecks by integrating Kafka event consumers.
- ▹Cut telecom onboarding latency by 90% (from days to hours) by automating A2P 10DLC registration and Twilio/Bandwidth profile provisioning via Rails.
- ▹Engineered an automated data ingestion pipeline syncing 500k+ district staff records daily across internal products using Nest.js and SFTP/CSV integrations.
- ▹Mentored a squad of 6 engineers across sprint cycles, establishing RFC design reviews and weekly 1:1 technical coaching.
- ▹Accelerated engineer onboarding and debugging by 50% by building a production-like Docker Compose environment simulating distributed services and Kafka brokers.
- ▹Elevated test coverage from 0% to 75% on legacy code and maintained 90%+ on new services through automated CI gates and pre-commit hooks.
- ▹Automated release workflows across 6 repositories using Conventional Commits, cutting weekly release overhead from 3 hours to zero manual intervention.
Senior Software Engineer
Nonprofit · Donations · Payments
- ▹Engineered a high-conversion, accessible donation widget using vanilla JavaScript and Web Components, reducing asset bundle size by 40% across mobile devices.
- ▹Optimized international payment processing by integrating Stripe and PayPal smart routing, reducing transaction failure rates by 15% across 30+ currencies.
- ▹Reduced CI suite execution time by 35% and eliminated intermittent test flakes by replacing Selenium with Cuprite for headless Chrome testing.
Senior Software Engineer
E-commerce · Marketplace
- ▹Reduced database read load by 60% and slashed p95 query latency to under 50ms by architecting Elasticsearch clusters for product search and catalog.
- ▹Established modular domain boundaries by architecting the company's first isolated Rails Engine, decoupling checkout logic and speeding up deploy cycles by 25%.
- ▹Eliminated 99% of order synchronization errors by building a resilient ETL ingestion pipeline handling high-volume external marketplace orders via Sidekiq.
- ▹Cut frontend asset build times by 70% in CI pipelines by executing migration from Webpacker to modern Rails Importmaps.
- ▹Accelerated GitLab CI feedback loops from 22 minutes to 8 minutes through test deduplication, database indexing, and parallel job orchestration.
- ▹Eliminated price race conditions for 50k+ daily concurrent shoppers by implementing real-time inventory broadcasts via ActionCable and Hotwire Turbo Streams.
Senior Ruby on Rails Engineer
Software Consultancy
- ▹Curacubby: Raised test suite coverage from 60% to 86% across billing systems, reducing production regression incidents by 40%.
- ▹Curacubby: Led migration of core application from Rails 4 to Rails 6, resolving security vulnerabilities and improving response times by 20%.
- ▹Curacubby: Halved CI pipeline runtimes by configuring parallel test execution, static analysis with RuboCop, and modular Rails Engines.
- ▹Kueski: Engineered a low-latency fraud-scoring microservice processing 10k+ daily loan applications with sub-100ms response times.
- ▹Kueski: Integrated risk evaluation APIs with core lending systems using strict OpenAPI contracts, cutting contract validation errors to zero.
Ruby on Rails Engineer
Software Consultancy
- ▹Promoted from Junior to Senior Engineer within 3 years, leading 4 client backend implementations across fintech and logistics domains.
- ▹Architected Prepaid2Cash transaction backend and integrated Sift Science fraud detection, reducing fraudulent transactions by 30% while processing millions in card volume.
- ▹Migrated legacy infrastructure from Rackspace to AWS and upgraded Rails 3 to Rails 4 for Orbridge, achieving 99.9% uptime and automated backups.
Personal Projects

KIDD🌀TS
Minimalistic dotfiles designed for productivity and performance. Featuring Kitty terminal, Zsh with Antidote, and a power-user Neovim setup.

DotFlakes
A reproducible NixOS configuration using Flakes and Home Manager. Prioritizes system stability with a polished GNOME desktop environment.
Writings & Insights
Deep dives into software architecture, patterns, AI engineering, and my daily learnings.

My AI Engineering Workflow: Skills, MCP Servers, and an LLM Wiki
I run AI as a small system, not a chat window. AGENTS.md holds the rules, an LLM wiki behind an MCP server holds the memory, and agent skills run the repeatable jobs. Here is how it works and what it cut.

Mastering the Repository Pattern in Ruby on Rails
Learn how to decouple your business logic from ActiveRecord and build more maintainable Rails applications using the Repository pattern.

Enforcing Local Quality Checks in Rails with Lefthook
Learn how to use Lefthook in Rails projects to enforce Rubocop, RSpec, Undercover, and migration checks locally — and why local hooks plus CI lead to healthier teams and more reliable codebases.

Avoiding N+1 Queries in Rails: Bullet, Prosopite, and Strict Loading
N+1 queries are one of the most common performance problems in Rails. In this post, I explain how to deal with them using Bullet, Prosopite, and the Rails strict loading feature.